Effects of background removal in GPR microwave tomography for archaeological and forensic studies

GPR method has been used worldwide in archaeological sites because of its fast data collecting and high-resolution results. However, the data collected usually present a high level of clutter caused by inhomogeneities in the background geologic material. The interpretation of the data may be a hard task depending on these clutter levels. In this framework, this work aims to evaluate the applicability of a background removal technique based on eigenvalue analysis in order to reduce the clutter level and emphasize the targets of interest. Analyses through a microwave tomographic imaging technique were performed before and after the background removal in order to evaluate the estimation of size and location of targets buried in controlled test sites. The results show that the clutter level is significantly reduced in the tomographic images after the background removal, when compared to the images obtained before this procedure. A slight reduction in the amplitude of the anomalies caused by the targets was also observed in the tomographic images; however, it was not enough to compromise the interpretation. When dealing with forensic applications the analysis must be performed carefully, since the clutter may be an important indicative of a burial location.